Open Access
Add to BibSonomyAbstract
The combined effects of a few mechanisms for emission efficiency enhancement produced in the overgrowth of the transparent conductor layer of Ga-doped ZnO (GaZnO) on a surface Ag-nanoparticle (NP) coated light-emitting diode (LED), including surface plasmon (SP) coupling, current spreading, light extraction, and contact resistivity reduction, are demonstrated. With a relatively higher GaZnO growth temperature (350 °C), melted Ag NPs can be used as catalyst for forming GaZnO nanoneedles (NNs) through the vapor-liquid-solid growth mode such that light extraction efficiency can be increased. Meanwhile, residual Ag NPs are buried in a simultaneously grown GaZnO layer for inducing SP coupling. With a relatively lower GaZnO growth temperature (250 °C), all the Ag NPs are preserved for generating a stronger SP coupling effect. By using a thin annealed GaZnO interlayer on p-GaN before Ag NP fabrication, the contact resistivity at the GaZnO/p-GaN interface and hence the overall device resistance can be reduced. Although the use of this interlayer blue-shifts the localized surface plasmon resonance peak of the fabricated Ag NPs from the quantum well emission wavelength of the current study (535 nm) such that the SP coupling effect becomes weaker, it is useful for enhancing the SP coupling effect in an LED with a shorter emission wavelength.
© 2015 Optical Society of America
1. Introduction
Although indium-tin-oxide (ITO) has been widely used as the transparent conductor in a light-emitting diode (LED) for the functions of current spreading and light extraction, searching for replacements is demanding because of the scarcity of indium on the earth. Among the materials under study, In-, Al-, and Ga-doped ZnO (GaZnO) have attracted much attention because of their high transparency in the visible range and high conductivity. In particular, due to its high thermal stability, GaZnO has been widely grown on lateral LEDs for improving their performances [1–13]. It has been shown that highly-doped GaZnO can lead to a resistivity level as low as 1.8 x 10−4 Ω-cm and transparency higher than 95% in the whole visible range [14]. On the other hand, surface nanostructures, particularly out-of-plane nanowires, have been fabricated on the p-GaN layer for enhancing light scattering and hence increasing light extraction efficiency [15–19]. In this regard, the commonly used approach is to grow ZnO nanorods on an LED based on the hydrothermal or other techniques [16–19]. Nanostructured ITO has also been fabricated on an LED for enhancing its light extraction [20–22]. Therefore, nanostructures of a new transparent conductor are also needed for further increasing the light extraction efficiency of an LED. Recently, this research team successfully grew highly-conductive GaZnO nanoneedles (NNs) based on the vapor-liquid-solid (VLS) growth mode with molecular beam epitaxy (MBE) by using Ag nanoparticles (NPs) as catalyst [23]. The high conductivity of the GaZnO NNs leads to extremely low threshold and turn-on electric fields in field emission tests. For a given Ag NP size, GaZnO NNs can be formed only when the growth temperature is high enough for melting at least part of an Ag NP. During the process of NN growth, the melted part of an Ag NP is pushed upward by the NN and becomes smaller and smaller due to the mixture of Ag atoms into the GaZnO NN. If such a growth temperature is not very high, residual Ag NPs of a lower density can be found around the NN bottom, which are covered by a simultaneously-grown GaZnO thin film. The residual Ag NPs include those originally larger Ag NPs, which cannot be melted at the GaZnO growth temperature for NN formation, and those partly melted and used for NN formation. Below the threshold growth temperature, which is between 250 and 350 °C when Ag NP size is at the order of a few tens nm, no GaZnO NN can be formed and the Ag NPs are buried in a GaZnO thin film. The resistivity of a GaZnO film can reach a low level of ~1.8 x 10−4 Ω-cm when it is grown at 250 °C. It increases as the growth temperature is raised. However, at the growth temperature of 350 °C, the resistivity can still be quite low at ~2.5 x 10−4 Ω-cm [14,24,25]. It is noted that because of the n-type nature of GaZnO, current injection into the p-GaN layer of an LED relies on a tunneling process. In particular, the work function of GaZnO (4.2-4.5 eV) [26] is significantly lower than that of p-GaN (7.6 eV) [27], leading to a high contact resistivity level between GaZnO and p-GaN. Therefore, with an as-grown GaZnO layer on p-GaN in an LED, the device resistance is usually higher than that with an ITO layer or a Ni/Au current spreading layer. Recently, thermal annealing processes under various conditions have been used for inducing the atomic inter-diffusion between GaZnO and p-GaN and hence reducing their contact resistivity [8,16,18,20–22].
Surface plasmon (SP) coupling with the quantum wells (QWs) in an LED has been proved to be useful for enhancing its internal quantum efficiency (IQE) [28–30], reducing its efficiency droop effect [30–32], and increasing its modulation bandwidth [33]. Among various metal nanostructures to induce SP resonance for coupling with the QWs, the formation of randomly-distributed surface metal NPs (Ag NPs for visible LEDs) represents one of the most inexpensive approaches [30,31,34,35]. Such an Ag NP distribution can be formed by simply thermally annealing a deposited Ag thin film. Usually, the spectral peak of the localized surface plasmon (LSP) resonance of surface Ag NPs on GaN is located in the green range. For the SP-coupling application to blue LEDs, a dielectric interlayer (DI) of a lower refractive index, when compared with that of GaN (~2.4), can be used between the Ag NPs and the p-GaN layer for blue-shifting the LSP resonance peak and hence enhancing the SP-coupling strength of a blue LED [30,31]. This blue-shift can also be used for compensating the red-shift of LSP resonance peak when the Ag NPs are covered by a transparent conductor, such as GaZnO, for enhancing current spreading and light extraction.
In this paper, we demonstrate the combination of several efficiency-enhancement mechanisms in green LEDs through a single surface-structure fabrication procedure of growing GaZnO NNs with the VLS mode. These mechanisms include the current spreading and light extraction of a GaZnO thin film, the light extraction of GaZnO NNs, the SP coupling through Ag NPs, the DI effect for blue-shifting LSP resonance peak, and the reduction of the contact resistivity between GaZnO and p-GaN through thermal annealing. Although each efficiency-enhancement mechanism has been previously reported, their combination through a single surface-structure fabrication procedure has never been reported. Two LED epitaxial structures of different p-type layer thicknesses are used for confirming the SP coupling effect in the structure of a thinner p-type layer. Two GaZnO growth temperatures are used for differentiating the effects of GaZnO NN formation. The LED samples without surface Ag NPs are also prepared for demonstrating the current spreading and light extraction effects of a thin GaZnO layer. Meanwhile, part of LED samples are fabricated with the formations of thermally annealed thin GaZnO layers (~10 nm) before Ag NP deposition for blue-shifting the LSP resonance peak and reducing the contact resistivity of the GaZnO/p-GaN interface. In section 2, the sample structures and their preparation procedures are presented. The basic optical characterization results are shown in section 3. Then, the comparisons of LED performance are reported in section 4. Discussions about the measured results are made in section 5. Finally, conclusions are drawn in section 6.
2. Sample structures and preparation procedures
In Table 1, we show the key structure features of the 20 LED samples prepared for comparing their performances. As shown in column 2, the LED samples are fabricated based on the epitaxial structures with the p-GaN layer thickness at either 50 or 120 nm. The two LED epitaxial structures are grown with metalorganic chemical vapor deposition on c-plane sapphire substrate. Either one consists of a 1.5-μm n-GaN layer, a five-period InGaN/GaN QW structure emitting light of ~535 nm in spectral peak, a 20-nm p-AlGaN layer, and a p-GaN layer of either 50 or 120 nm in thickness. With the p-GaN thickness at 120 nm, the distance between the surface Ag NPs and the top QW is as large as ~150 nm (including the 20-nm p-AlGaN layer and a 10-nm QW-barrier) such that SP coupling is weak. As shown in Table 1, the two reference samples R-1 and R’-1 represent the standard lateral LEDs with metal current-spreading layers of Ni/Au (5/5 nm) on the p-GaN layers. On the p-GaN layers of the other two reference samples R-2 and R’-2, surface Ag NPs are formed before the deposition of the metal current-spreading layers of Ni/Au (5/5 nm). The Ag NPs are formed by depositing an Ag layer of ~2 nm, followed by a process of thermal annealing at 200 °C for 30 min. No GaZnO is grown on these four reference samples. On the p-GaN layers of samples A-350 and A’-350 (A-250 and A’-250), after the similar Ag NPs are formed, GaZnO is grown for 80 min with a RF plasma-assisted MBE reactor at 350 (250) °C in substrate temperature under the Zn-rich growth conditions of 320 °C in Zn effusion cell temperature, 900 °C in Ga effusion cell temperature, 1 sccm in O2 flow rate, and 350 W in RF-plasma power. On the p-GaN layers of samples B-350 and B’-350 (B-250 and B’-250), GaZnO is grown under the same conditions as those for samples A-350 and A’-350 (A-250 and A’-250). However, no Ag NP is formed on p-GaN before GaZnO growth. Figures 1(a) and 1(b) [1(c) and 1(d)] show the plan-view scanning electron microscopy (SEM) images of the top surface before and after GaZnO growth, respectively, of sample A-350 (A-250). From either Fig. 1(a) or 1(c), we can estimate the average size and planar density of Ag NP on p-GaN to give 21.6 nm and 1.03 x 1011 cm−2, respectively. Here, one can see that when GaZnO is grown at 350 °C, NNs are formed. No NN can be seen on the top surface when GaZnO is grown at 250 °C. It is noted that the orientations of the formed GaZnO NNs are quite random. The GaZnO NNs with random orientations shown in Fig. 1(b) are quite different from those shown in [23], which are parallel and vertical. The cause for such a difference will be discussed in section 5. The surface morphologies before and after GaZnO growth of sample A’-350 (A’-250) are similar to those of sample A-350 (A-250). Because no Ag NP is fabricated on p-GaN in samples B-350, B’-350, B-250, and B’-250, no GaZnO NN is formed on the tops of these samples. It is noted that a GaZnO thin film is simultaneously deposited when GaZnO NNs are formed. The thickness of the overgrown GaZnO layer is ~180 (~250) nm when NNs are (no NN is) formed.
Table 1. Key Structure Features of the 20 LED Samples Prepared for Comparing their Performances.
Fig. 1 (a) and (b) [(c) and (d)]: Plan-view SEM images of the top surface before and after GaZnO growth, respectively, of sample A-350 (A-250).
To blue-shift the LSP resonance peak and to reduce the contact resistivity at the GaZnO/p-GaN interface, we prepare another series of sample with a 10-nm GaZnO layer grown before Ag NP formation. This 10-nm GaZnO layer is grown at 250 °C and then thermally annealed at 600 °C for 15 min in the MBE chamber. This thin annealed-GaZnO layer between p-GaN and subsequently formed Ag NPs can serve as a DI for blue-shifting the LSP resonance peak of the Ag NPs and also as a buffer layer (annealed junction) for reducing the contact resistivity at the GaZnO/p-GaN interface. The annealing temperature at 600 °C is an optimized choice. As shown in Table 1, samples C-350, C’-350, D-350, D’-350, C-250, C’-250, D-250, and D’-250 are prepared under the same conditions as those of samples A-350, A’-350, B-350, B’-350, A-250, A’-250, B-250, and B’-250, respectively, except the annealed 10-nm GaZnO interlayer grown before Ag NP formation. Figures 2(a) and 2(b) [2(c) and 2(d)] show the plan-view SEM images of the top surface before and after GaZnO growth, respectively, of sample C-350 (C-250). From either Fig. 2(a) or 2(c), we can estimate the average size and planar density of Ag NP on a 10-nm GaZnO layer to give 12.3 nm and 2.56 x 1011 cm−2, respectively. Here, one can also see that when GaZnO is grown at 350 °C, NNs are formed. No NN can be observed on the top surface when GaZnO is grown at 250 °C. It is noted that under the same preparation condition, the Ag NP sizes formed on the 10-nm GaZnO interlayer are smaller than those formed directly on p-GaN. The smaller Ag NPs in Fig. 2(a) lead to thinner NNs in Fig. 2(b). The formation of the smaller Ag NPs shown in Figs. 2(a) and 2(c) can be due to the rougher surface on the 10-nm GaZnO interlayer.
Fig. 2 (a) and (b) [(c) and (d)]: Plan-view SEM images of the top surface before and after GaZnO growth, respectively, of sample C-350 (C-250).
Figures 3(a) and 3(b) show the cross-sectional transmission electron microscopy (TEM) images near the GaZnO/p-GaN interfaces of samples A-350 and C-250, respectively. In Fig. 3(a) for sample A-350, the darker body labelled by “NN” corresponds to the “root” of a GaZnO NN, which is the portion of an NN covered by the simultaneously grown GaZnO film [23]. Around the GaZnO/p-GaN interface, a few dark spots exist corresponding to the residual Ag NPs, which are not used for NN growth. In Fig. 3(b) for sample C-250, a denser distribution of Ag NP can be seen in the GaZnO layer at a distance of ~10 nm from the GaZnO/p-GaN interface. The Ag NPs in this sample are all preserved because no GaZnO NN is formed at this growth temperature (250 °C). With more buried Ag NPs in sample C-250 (also in sample A-250), SP-coupling is expected to be stronger in this sample, when compared with sample A-350 (also sample C-350).
Fig. 3 (a) and (b): Cross-sectional TEM images near the GaZnO/p-GaN interfaces of samples A-350 and C-250, respectively. In part (a), the darker body labeled by “NN” corresponds to the “root” of a GaZnO NN. The dark spots correspond to the preserved Ag NPs with some of them indicated by the thicker arrows.
Figure 4(a) schematically demonstrate the general structure of the fabricated LED devices (except the four reference samples). Only in samples A-350, A’-350, C-350, and C’-350, GaZnO NNs are formed at the top. In these samples, residual Ag NPs may exist near the GaZnO/p-GaN interface. In samples A-250, A’-250, C-250, and C’-250, all the prepared Ag NPs are preserved to produce strong LSP resonances. In samples B-350, B’-350, B-250, B’-250, D-350, D’-350, D-250, and D’-250, neither Ag NP nor GaZnO NN exists in the devices. Only a GaZnO layer is deposited at the top of each of those samples. The LED devices are fabricated following the standard procedures with a square mesa of 300 μm x 300 μm in size. The plan-view SEM image of an LED device of sample A-350 is shown in Fig. 4(b). Here, the brighter central circular area and square-grid region correspond to the p-contact, which is formed by depositing Ti of 20 nm and then Au of 200 nm in thickness. No metal deposition is applied outside the p-contact area on the mesa. An area around the vertical arrow in Fig. 4(b) is magnified to give the image in Fig. 4(c). Here, one can see that the GaZnO NNs are well preserved in the region outside the p-contact (the left portion) after the whole device process. In the p-contact region, the NNs are not broken even though metal layers are deposited on them. The n-contact of the device is formed outside the mesa [not shown in Fig. 4(b)] by also depositing Ti of 20 nm and then Au of 200 nm in thickness. It is noted that the p-contacts of the four reference samples (R-1, R’-1, R-2, and R’-2) consist of the metal layers of Ni/Au of 20/200 nm in thickness.
Fig. 4 (a): Schematic demonstration of the general structure of the fabricated LED devices (except the four reference samples). (b): Plan-view SEM image of an LED device of sample A-350. (c): Magnified SEM image in an area around the vertical arrow in part (b).
3. Optical characterization results
Figure 5(a) shows the transmission spectra of samples A-350 and A’-350 before and after GaZnO growth. For comparison, the transmission spectra of samples R-2 and R’-2 are also shown in this figure. The vertical dashed line here roughly indicates the QW emission wavelength (~535 nm). Before GaZnO growth in samples A-350 and A’-350 (curves labeled by “A-350/NP” and “A’-350/NP”, respectively) and in samples R-2 and R’-2, the clear depressions in the transmission spectra are caused by the absorption and scattering induced by LSP resonance. The depression depth at a given wavelength corresponds to the LSP resonance strength and hence SP coupling strength at this wavelength. In Fig. 5(a), the LSP resonance features of samples A-350 and A’-350 before GaZnO growth and samples R-2 and R’-2 are similar, indicating that the morphologies of Ag NPs on those samples are about the same. After GaZnO growth, the transmission spectra of samples A-350 and A’-350 change dramatically. Here, the generally decreasing transmission with decreasing wavelength is caused by the Rayleigh-like scattering of the grown GaZnO NNs [36]. Local depressions with the minima around 550 nm can still be observed, indicating that relatively weaker LSP resonance features are induced through the residual Ag NPs after GaZnO growth (forming NNs) in samples A-350 and A’-350. To understand the depths of LSP-resonance induced transmission depressions in samples A-350 and A’-350, we use the curves of a + bλ−4 (λ is wavelength, a and b are two fitting constants) for fitting the curves of samples A-350 and A’-350 by matching the transmittance levels as possible at 400 and 750 nm. The two fitting curves are labeled by “RS” in Fig. 5(a) and correspond to the effect of background Rayleigh-like scattering by the grown GaZnO NNs. The difference between the curve of sample A-350 or A’-350 and the corresponding fitting curve shows the effect of LSP resonance induced by the residual Ag NPs. By setting the reference level of this LSP-resonance induced transmission spectrum at 750 nm as 100%, we can plot the curves labeled by A-350/SP and A’-350/SP in Fig. 5(a). These two curves can be used for understanding the LSP resonance strengths of the residual Ag NPs in samples A-350 and A’-350, respectively. Similar transmission spectra of samples A-250 and A’-250 are shown in Fig. 5(b). Here, the transmission spectra of samples A-250 and A’-250 before GaZnO growth are slightly red-shifted from those of samples R-2 and R’-2, indicating the slightly larger Ag NP sizes formed in samples A-250 and A’-250, when compared with those in samples A-350 and A’-350. However, such a difference does not affect the conclusions drawn in this paper. After GaZnO growth in samples A-250 and A’-250, the deeper depressions indicate the strong LSP resonances induced by the preserved Ag NPs. Because they are buried in GaZnO with a refractive index around 1.8, the LSP resonance peaks are red-shifted to 525.6 and 524.2 nm in samples A-250 and A’-250, respectively. Transmission spectra of samples C-350 and C’-350 (samples C-250 and C’-250), similar to those of samples A-350 and A’-350 (samples A-250 and A’-250) in Fig. 5(a), are shown in Fig. 5(c) [Fig. 5(d)]. As shown in Figs. 1(a), 1(c), 2(a), and 2(c), the sizes of the fabricated Ag NPs on the annealed 10-nm GaZnO interlayer are smaller than those fabricated directly on p-GaN. Therefore, the depressions of curves C-350/NP and C’-350/NP (C-250/NP and C’-250/NP) can be blue-shifted from those of curves R-2 and R’-2 in Fig. 5(c) [Fig. 5(d)]. However, the major reason for the blue shifts of the depressions of curves C-350/NP, C’-350/NP, C-250/NP, and C’-250/NP from those of samples R-2 and R’-2 is the use of the annealed 10-nm GaZnO interlayer. Also, with the smaller Ag NPs, almost all the Ag NPs are completely melted at 350 °C for GaZnO NN formation such that after GaZnO growth at 350 °C, only a few residual Ag NPs are left for inducing an LSP resonance depression feature in either transmission spectrum of C-350 or C’-350 in Fig. 5(c). The fitting curves of a + bλ−4 for the curves of samples C-350 and C’-350 similar to those for the curves of samples A-350 and A’-350 in Fig. 5(a) are plotted (curves labeled by “RS”) in Fig. 5(c). Also, the curves labeled by “C-350/SP” and “C’-350/SP” show the differences between the fitting curves and the curves of C-350 and C’-350, respectively, for demonstrating the LSP resonance strengths of the residual Ag NPs in these two samples. Similar to samples A-250 and A’-250 in Fig. 5(b), the transmission curves of samples C-250 and C’-250 after GaZnO growth in Fig. 5(d) show red-shifted depressions induced by the preserved Ag NPs buried in GaZnO. Because of the use of the 10-nm GaZnO interlayers in samples C-250 and C’-250, their depression-minimum wavelengths are shorter, when compared with those of samples A-250 and A’-250. The transmission-minimum wavelengths of those samples with fabricated Ag NPs are listed in column 2 of Table 2. Also, the transmission depression depths at the QW emission wavelength (535 nm) of various samples are listed in column 3 of Table 2. Except curves A-350/SP, A’-350/SP, C-350/SP, and C’-350/SP, which are used for evaluating the LSP resonance strengths in samples A-350, A’-350, C-350, and C’-350, the transmission depression depths are defined with respect to the individual levels at 750 nm in all samples. In curves A-350/SP, A’-350/SP, C-350/SP, and C’-350/SP, the depression depths are defined with respect to 100%. Normally, a lower transmittance level corresponds to stronger LSP resonance and hence a stronger SP-coupling effect. Therefore, we expect strong SP coupling effects in samples R-2, R’-2, A-250, A’-250, C-250, and C’-250. The SP coupling strengths in samples A-350 and A’-350 are expected to be relatively weaker. Those in samples C-350 and C’-350 are supposed to be weak.
Fig. 5 (a): Transmission spectra of samples A-350 and A’-350 before and after GaZnO growth. Before GaZnO growth, the curves are labeled by “A-350/NP” and “A’-350/NP”. The fitting curves of Rayleigh-like scattering are labeled by “RS”. The curves for the calibrated LSP resonance depressions are labeled by “A-350/SP” and “A’-350/SP”. (b): Similar transmission spectra of samples A-250 and A’-250. (c): Similar transmission spectra of samples C-350 and C’-350. (d): Similar transmission spectra of samples C-250 and C’-250. For comparison, the transmission spectra of samples R-2 and R’-2 are also shown in these figures. The vertical dashed lines roughly indicate the QW emission wavelength (~535 nm).
Table 2. Characterization Results of the 20 LED Samples, Including Transmission Data, IQE, Device Resistance, Normalized Output Intensity, and Efficiency Droop Range
Temperature-dependence photoluminescence (PL) measurements are performed for all the 20 LED samples to calibrate their IQEs. Figure 6 shows the variations of integrated PL intensity with temperature of samples A-350, A-250, C-350, C-250, R-1, R’-1, R-2, and R’-2. An IQE is defined as the ratio of integrated PL intensity at 300 K over that at 10 K. The IQEs of all the samples are listed in column 4 of Table 2 (the numbers before slashes). The numbers after slashes represent the ratios of the IQEs of those samples of 50-nm (120-nm) p-GaN with respect to that of R-1 (R’-1). Here, one can see the generally higher IQEs in the samples fabricated with the epitaxial structure of 120-nm p-GaN, when compared with those fabricated with the epitaxial structure of 50-nm p-GaN. This difference is due to the more effective thermal annealing during the growth of thicker p-GaN in the epitaxial structure with 120-nm p-GaN [37]. By comparing the IQEs between samples R-1 and R-2, one can see that the SP coupling induced by the surface Ag NPs increases IQE from 20.4 to 28.1%. The insignificant increase of IQE from 27.6% of sample R’-1 to 28.8% of sample R’-2 indicates that the SP coupling effect is weak in the samples with 120-nm p-GaN. The IQEs of all the samples with 120-nm p-GaN are about the same within the range of 27.2-28.9%. Among the samples with 50-nm p-GaN, the IQEs of those samples without a significant SP coupling effect (including sample C-350) are maintained at the similar levels to that of sample R-1 (20.4-21.2%). The IQEs of those samples with strong SP coupling, including R-2, A-350, A-250, and C-250, are significantly increased.
Fig. 6 Integrated PL intensity variations with temperature of samples A-350, A-250, C-350, C-250, R-1, R’-1, R-2, and R’-2.
4. Comparison of LED performances
Figures 7(a) shows the relations between injected current and applied voltage (I-V curves) of samples A-350, A’-350, B-350, B’-350, R-1, R’-1, R-2, and R’-2. Its inset shows the magnified portion in the voltage range of 4.6-5.4 V for differentiating those curves. Similar I-V curves of samples A-250, A’-250, B-250, and B’-250 are shown in Fig. 7(b) and its inset. Also, similar I-V curves of samples C-350, C’-350, D-350, and D’-350 (C-250, C’-250, D-250, and D’-250) are shown in Fig. 7(c) [Fig. 7(d)] and its inset. Here, one can see that the turn-on voltages of all the 20 LED samples are around 3 V. Also, no significant leakage current is observed up to 10 V under reverse bias. The device resistance levels of all those LED samples are listed in column 5 of Table 2. From the resistance data, a few variation trends can be observed. First, the device resistance levels of those samples with 120-nm p-GaN are lower than those of their corresponding samples with 50-nm p-GaN. This is so because of the more effective current spreading in the samples with thicker p-GaN layers. Second, with an annealed 10-nm GaZnO interlayer on p-GaN for reducing the contact resistivity (in samples C-350, C’-350, D-350, D’-350, C-250, C’-250, D-250, and D’-250), the device resistance levels are lower, when compared to those of the corresponding samples without such an annealed GaZnO layer (in samples A-350, A’-350, B-350, B’-350, A-250, A’-250, B-250, and B’-250). In particular, the device resistance levels of samples C-250, C’-250, D-250, and D’-250, in which the GaZnO layers are grown at 250 °C, are lower than those of the corresponding reference samples (R1, R’-1, R-2, and R’-2). Third, the device resistance levels of those samples with their GaZnO layers grown at 250 °C are lower than those of the corresponding samples grown at 350 °C. This is so because the GaZnO conductivity is higher when it is grown at 250 °C. Fourth, the device resistance levels of those samples with Ag NPs (with or without GaZnO NN formation) are generally higher than those of the samples without Ag NP. The distribution of Ag NPs around the GaZnO/p-GaN interface can block current flow and slightly increase device resistance. Fifth, the device resistance levels of those samples with GaZnO NN formation are higher than those of the samples without GaZnO NN formation. This is so because we fabricate p-contacts directly on the NNs such that electrical current flows through the NNs. Although the conductivity of an NN can be quite high, the current flow cross-section through NNs is smaller than that of a solid GaZnO layer and hence the device resistance is higher.
Fig. 7 (a): I-V curves of samples A-350, A’-350, B-350, B’-350, R-1, R’-1, R-2, and R’-2. The inset shows the magnified portion in the voltage range of 4.6-5.4 V. (b): Similar I-V curves of samples A-250, A’-250, B-250, and B’-250. (c): Similar I-V curves of samples C-350, C’-350, D-350, and D’-350. (d): Similar I-V curves of samples C-250, C’-250, D-250, and D’-250.
Figures 8(a)-8(d) show the variations of normalized output intensity with injected current (L-I curves) for those samples in Figs. 7(a)-7(d), respectively. All the LED output intensities are normalized with respect to the level of sample R-1 at 100 mA in injected current. The LED output intensity in each sample is obtained by measuring the output intensities from the top and bottom sides of the device and taking the summation. The normalized output intensities of all the 20 LED samples at 100 mA in injected current are listed in column 6 of Table 2 (the numbers before slashes for those samples with 120-nm p-GaN). The numbers after slashes for those samples with 120-nm p-GaN represent their output intensities normalized with respect to the level of sample R’-1 at 100 mA in injected current. Therefore, those numbers after slashes can be used for the comparisons among the samples based on the epitaxial structure with a 120-nm p-GaN layer. From the output intensity results shown in Table 2, we can observe the following variation trends. (1) The output intensities of the samples fabricated with the epitaxial structure of 120-nm p-GaN are higher, when compared to those of the corresponding samples fabricated with the epitaxial structure of 50-nm p-GaN, except the comparison between samples A-250 and A’-250. This result is attributed to two factors, including the lower device resistance leading to a lower heating effect and the higher QW IQE in the samples with 120-nm p-GaN. However, the normalized intensities with respect to the levels of the individual reference samples (R-1 and R’-1, respectively) are generally higher in the samples with 50-nm p-GaN, particularly for those samples with significant SP coupling. The large distance between the Ag NPs and the top QW (~150 nm) in the samples with 120-nm p-GaN results in weak SP coupling even though the fabricated Ag NPs produce strong LSP resonances, as shown in Figs. 5(a), 5(b), and 5(d). (2) Among the samples with 50-nm p-GaN, sample A-250 has the highest output intensity, followed by sample C-250, A-350, and then C-350. The factors for emission efficiency enhancement in sample A-250 include the embedded Ag NPs for SP coupling and the high-conductivity GaZnO thin film (grown at 250 °C) for current spreading and light extraction. Nevertheless, the key factor is the strong SP coupling due to the strong LSP resonance and the coincidence of the LSP resonance peak with the QW emission wavelength [see Fig. 5(b)]. Although the LSP resonance is quite strong in sample C-250, its peak is away from the QW emission wavelength such that the SP coupling effect is not so strong, leading to relatively lower output intensity. The blue-shifted LSP resonance peak with respect to the QW emission wavelength in samples C-250 is mainly caused by the use the 10-nm GaZnO interlayer for blue-shifting the LSP resonance peak. Therefore, although the annealed GaZnO interlayer can help in reducing device resistance, its contribution to effective SP coupling depends on the QW emission wavelength. Then, in sample A-350, the formed GaZnO NNs are expected to enhance light extraction and increase output intensity. Because the LSP resonance and hence SP coupling are relatively weaker in sample A-350, when compared to samples C-250, the comparable output intensities of the two samples indicate the importance of the light extraction effect of GaZnO NNs. However, the significant lower output intensity in sample C-350, when compared with sample A-350, shows that the SP coupling caused by the residual Ag NPs is still a crucial factor for emission enhancement. (3) Among those samples without Ag NP fabrication, the output intensities in either group of sample with 50- or 120-nm p-GaN when they are normalized with respect to the levels of individual reference samples are about the same, in the range of 1.45-1.53. The enhancements are attributed to the current spreading and light extraction effects of the overgrown GaZnO layers. With the annealed 10-nm GaZnO interlayer, the lower device resistance can slightly increase output intensity. The systematically slightly higher output intensities in those samples with the GaZnO growth temperature at 350 °C, when compared with those with the GaZnO growth temperature at 250 °C, can be due to the rougher GaZnO surface grown at 350 °C for stronger light extraction even though the GaZnO conductivity grown at 250 °C is higher. The GaZnO surface roughness levels are 4.96 and 1.99 nm when the GaZnO layers are grown at 350 and 250 °C, respectively.
Fig. 8 (a)-(d): Variations of normalized output intensity with injected current for the samples in Figs. 7(a)-7(d), respectively.
Figures 9(a)-9(d) show the relative efficiency variations with injected current for those samples in Figs. 8(a)-8(d), respectively. The relative efficiency is defined as the ratio of output optical power over the input electrical power and then normalized with respect to the highest level among the samples in the same figure. The droop ranges of all the 20 LED samples are listed in column 7 of Table 2. The droop range of a sample is defined as the efficiency reduction percentage from the individual maximum efficiency to its level at 100 mA in injected current. From Table 2, one can see that generally the droop range decreases with increasing output intensity or emission efficiency. In particular, in those samples with strong SP coupling (samples A-250 and C-250), the efficiency droop ranges are significantly reduced.
Fig. 9 (a)-(d): Relative efficiency variations with injected current for the samples in Figs. 8(a)-8(d), respectively.
5. Discussions
The comparison of the variation trends among the studied samples between IQE and LED output intensity can help in differentiating the effects of SP coupling and light extraction through GaZnO films and NNs on emission efficiency enhancement. SP coupling can increase the IQE and hence output intensity of an LED. Enhanced light extraction can increase the output intensity of an LED, but not its IQE. By comparing the IQEs and output intensities of samples C-350 and R-1, the IQE of sample C-350 is increased by only ~4%. However, its output intensity increases by 69%, indicating the important contribution of the light extraction effects of GaZnO thin film and NNs. Similar comparison can be made between samples R-1 and all other samples of 50-nm p-GaN without Ag NP formation for showing the important contribution of the light extraction effect of a GaZnO thin film. By comparing samples A’-350 and C’-350 with all other samples of 120-nm p-GaN without the formation of GaZnO NNs, we can see that GaZnO NNs can indeed further increase light extraction beyond the effect of a GaZnO thin film. Among samples A-350, A-250, and C-250, which have significant SP coupling effects, the IQE values can indicate the SP coupling strengths. Although there are GaZnO NNs in sample A-350 for enhancing light extraction, the SP coupling effect is weaker in this sample, when compared with samples A-250 and C-250, and hence its output intensity is lower than that of sample A-250 and comparable to that of sample C-250. From column 7 of Table 2, one can see that besides SP coupling, higher intrinsic QW IQE and higher light extraction efficiency can reduce the efficiency droop effect.
The GaZnO NNs formed on the tops of samples A-350 and C-350, as shown in Figs. 1(b) and 2(b), respectively, have random orientations. They are quite different from the vertical and parallel GaZnO NNs published previously by this research group [23]. Such a difference in NN orientation is caused by the difference in Ag NP density. When the Ag NP density is relatively lower and the separation between neighboring Ag NPs is relatively larger, the growths of individual NNs are independent from each other. In this situation, most NNs are grown vertically. However, when the Ag NP density is high and the separation between neighboring Ag NPs is small, the growth of an NN can be influenced by those of the neighboring ones. In this situation, the precipitated GaZnO from an Ag NP close to another Ag NP nearby can be used as the seed for GaZnO precipitation of the second Ag NP and “pulls” its precipitation toward the first Ag NP such that the NN growth is bent. The planar densities of the formed Ag NPs in this study are generally higher than those for growing the NNs in [23] such that the GaZnO NNs grown in the current study have random orientations. The difference in light extraction effect between GaZnO NNs of random and regular orientations is an issue deserving further investigation.
The results above show that both SP coupling and GaZnO NN scattering can enhance the overall light emission efficiency of an LED. However, with effective NN formation, the residual Ag NP density is reduced and hence the SP coupling effect is weakened. One possible method to simultaneously achieve NN formation for enhancing light extraction and a high residual Ag NP density for strong SP coupling is the fabrication of a surface Ag NP distribution of a large size variety. In this situation, the smaller Ag NPs are used for growing GaZnO NNs and the larger Ag NPs are preserved for inducing SP coupling. If the preserved Ag NP density can be higher for increasing the LSP resonance strength around the QW emission wavelength at 535 nm [see Fig. 5(a)], a stronger SP coupling effect in sample A-350 can further increase its output intensity. The use of the annealed GaZnO interlayer before Ag NP fabrication can reduce the contact resistivity and hence the overall device resistance. However, the use of this layer blue-shifts the LSP resonance peak with the blue-shift range determined by the layer thickness. For a given QW emission wavelength, the GaZnO interlayer thickness and Ag NP size need to be well controlled for maximizing the LSP resonance strength at the QW emission wavelength after the growth of a GaZnO thin film. Nevertheless, based on the approaches described above, it is difficult to shift the LSP resonance peak to the blue range. In this situation, the SP coupling in a blue LED relies on the spectral shoulder of LSP resonance peak. Because SiO2 has a refractive index (~1.5) lower than that of GaZnO, the deposition of SiO2-coated Ag NPs on an LED can help in further blue-shifting the LSP resonance peak [38]. However, the use of this approach may increase LED fabrication cost. In the current study, although the SP coupling effects in those samples with 120-nm p-GaN are weak such that their efficiency enhancement ratios are not as large as those with 50-nm p-GaN, their absolute emission efficiencies are generally higher (except the comparison between samples A-250 and A’250) due to their higher intrinsic QW IQEs. Therefore, efforts are needed to improve the LED growth quality for achieving a higher QW IQE under the condition of a thin p-GaN layer for effective SP coupling.
6. Conclusions
In summary, we have demonstrated the combined effects of a few mechanisms for emission efficiency enhancement generated in the overgrowth of a transparent conductor layer of GaZnO on a surface Ag-NP coated LED, including SP coupling, current spreading, light extraction, and contact resistivity reduction. With a relatively higher GaZnO growth temperature (350 °C), Ag NPs could be used as catalyst for forming GaZnO NNs such that light extraction efficiency could be increased. In this situation, certain residual Ag NPs were buried in the simultaneously grown GaZnO layer for inducing SP coupling. With a relatively lower GaZnO growth temperature (250 °C), all the Ag NPs were preserved for generating a stronger SP coupling effect. By using a thin annealed GaZnO interlayer on p-GaN before Ag NP fabrication, the contact resistivity at the GaZnO/p-GaN interface and hence the overall device resistance could be reduced. Although the use of this GaZnO interlayer blue-shifted the LSP resonance peak of the fabricated Ag NPs from the QW emission wavelength (535 nm) in the current study such that the SP coupling effect became weaker, it is useful for enhancing the SP coupling effect in an LED with a shorter emission wavelength. From the illustrated results, one can see that SP coupling is the most effective mechanism for enhancing LED emission efficiency if the planar density of Ag NP is sufficiently high. However, scattering of dense NNs is also important for improving light extraction. The residual Ag NP density relies on the GaZnO growth condition. A higher growth temperature leads to a higher NN density, but a lower residual Ag NP density. In the future work, an optimized growth temperature between 250 and 350 °C can be identified for maximizing both grown NN density and residual Ag NP density such that both effects of NN scattering and SP coupling can be maximized for the most effective enhancement of LED overall efficiency.
Acknowledgments
This research was supported by Ministry of Science and Technology, Taiwan, under the grants of MOST 103-2120-M-002-002, NSC 102-2221-E-002-204-MY3, and MOST 103-2221-E-002-139, by National Taiwan University (103R890951 and 103R890952), and by US Air Force Scientific Research Office under the contract of AOARD-14-4105.
References and links
1. H. J. Ko, Y. F. Chen, S. K. Hong, H. Wenisch, T. Yao, and D. C. Look, “Ga-doped ZnO films grown on GaN templates by plasma-assisted molecular-beam epitaxy,” Appl. Phys. Lett. 77(23), 3761–3763 (2000). [CrossRef]
2. J. K. Sheu, Y. S. Lu, M. L. Lee, W. C. Lai, C. H. Kuo, and C. J. Tun, “Enhanced efficiency of GaN-based light-emitting diodes with periodic textured Ga-doped ZnO transparent contact layer,” Appl. Phys. Lett. 90(26), 263511 (2007). [CrossRef]
3. J. K. Sheu, M. L. Lee, Y. S. Lu, and K. W. Shu, “Ga-doped ZnO transparent conductive oxide films applied to GaN-based light-emitting diodes for improving light extraction efficiency,” IEEE J. Quantum Electron. 44(12), 1211–1218 (2008). [CrossRef]
4. Z. Liu, X. Wang, H. Yang, Y. Duan, and Y. Zeng, “A Ga-doped ZnO transparent conduct layer for GaN-based LEDs,” J. Semiconductors 31(9), 094002 (2010). [CrossRef]
5. J. H. Lim, K. H. Lee, and D. C. Lim, “Enhanced performance in GaN light emitting diode by patterned ZnO transparent conducting oxide,” J. Korean Phys. Soc. 57(5), 1229–1232 (2010). [CrossRef]
6. T. Y. Park, Y. S. Choi, J. W. Kang, J. H. Jeong, S. J. Park, D. M. Jeon, J. W. Kim, and Y. C. Kim, “Enhanced optical power and low forward voltage of GaN-based light-emitting diodes with Ga-doped ZnO transparent conducting layer,” Appl. Phys. Lett. 96(5), 051124 (2010). [CrossRef]
7. H. Y. Liu, X. Li, S. Liu, X. Ni, M. Wu, V. Avrutin, N. Izyumskaya, Ü. Özgür, A. B. Yankovich, A. V. Kvit, P. M. Voyles, and H. Morkoç, “InGaN based light emitting diodes utilizing Ga doped ZnO as a highly transparent contact to p-GaN,” Phys. Status Solidi, C Conf. Crit. Rev. 8(5), 1548–1551 (2011). [CrossRef]
8. K. Y. Yen, C. H. Chiu, C. W. Li, C. H. Chou, P. S. Lin, T. P. Chen, T. Y. Lin, and J. R. Gong, “Performance of InGaN/GaN MQW LEDs using Ga-doped ZnO TCLs prepared by ALD,” IEEE Photonics Technol. Lett. 24(23), 2105–2108 (2012). [CrossRef]
9. W. L. Bi, C. L. Ho, and M. C. Wu, “Ga-doped ZnO grown by atomic layer deposition and the application to blue light-emitting diodes as a current spreading layer,” ECS Solid State Lett. 2(11), Q98–Q100 (2013). [CrossRef]
10. R. H. Horng, K. C. Shen, C. Y. Yin, C. Y. Huang, and D. S. Wuu, “High performance of Ga-doped ZnO transparent conductive layers using MOCVD for GaN LED applications,” Opt. Express 21(12), 14452–14457 (2013). [CrossRef] [PubMed]
11. H. S. Shin, J. H. Lee, J. S. Kwak, H. H. Lee, and H. K. Kim, “Linear facing target sputtering of the epitaxial Ga-doped ZnO transparent contact layer on GaN-based light-emitting diodes,” J. Phys. D Appl. Phys. 46(41), 415301 (2013). [CrossRef]
12. K. Y. Yen, C. H. Chiu, C. Y. Hsiao, C. W. Li, C. H. Chou, K. Y. Lo, T. P. Chen, C. H. Lin, T. Y. Lin, and J. R. Gong, “Characteristics of GaN-based LEDs using Ga-doped or In-doped ZnO transparent conductive layers grown by atomic layer deposition,” J. Cryst. Growth 387, 91–95 (2014). [CrossRef]
13. M. S. Oh and I. Seo, “Enhanced performance of GaN-based green light-emitting diodes with gallium-doped ZnO transparent conducting oxide,” J. Electron. Mater. 43(4), 1232–1236 (2014). [CrossRef]
14. H. S. Chen, Y. F. Yao, C. H. Liao, C. G. Tu, C. Y. Su, W. M. Chang, Y. W. Kiang, and C. C. Yang, “Light-emitting device with regularly patterned growth of an InGaN/GaN quantum-well nanorod light-emitting diode array,” Opt. Lett. 38(17), 3370–3373 (2013). [CrossRef] [PubMed]
15. C. H. Lin, C. G. Tu, H. S. Chen, C. Hsieh, C. Y. Chen, C. H. Liao, Y. W. Kiang, and C. C. Yang, “Vertical light-emitting diodes with surface gratings and rough surfaces for effective light extraction,” Opt. Express 21(15), 17686–17694 (2013). [CrossRef] [PubMed]
16. H. Park, K. J. Byeon, K. Y. Yang, J. Y. Cho, and H. Lee, “The fabrication of a patterned ZnO nanorod array for high brightness LEDs,” Nanotechnology 21(35), 355304 (2010). [CrossRef] [PubMed]
17. K. S. Kim, S. M. Kim, H. Jeong, M. S. Jeong, and G. Y. Jung, “Enhancement of light extraction through the wave-guiding effect of ZnO sub-microrods in InGaN blue light-emitting diodes,” Adv. Funct. Mater. 20(7), 1076–1082 (2010). [CrossRef]
18. K. Dai, C. B. Soh, S. J. Chua, L. Wang, and D. Huang, “Influence of the alignment of ZnO nanorod arrays on light extraction enhancement of GaN-based light-emitting diodes,” J. Appl. Phys. 109(8), 083110 (2011). [CrossRef]
19. H. Jeong, Y. H. Kim, T. H. Seo, H. S. Lee, J. S. Kim, E. K. Suh, and M. S. Jeong, “Enhancement of light output power in GaN-based light-emitting diodes using hydrothermally grown ZnO micro-walls,” Opt. Express 20(10), 10597–10604 (2012). [CrossRef] [PubMed]
20. C. O’Dwyer, M. Szachowicz, G. Visimberga, V. Lavayen, S. B. Newcomb, and C. M. S. Torres, “Bottom-up growth of fully transparent contact layers of indium tin oxide nanowires for light-emitting devices,” Nat. Nanotechnol. 4(4), 239–244 (2009). [CrossRef] [PubMed]
21. C. H. Chiu, P. Yu, C. H. Chang, C. S. Yang, M. H. Hsu, H. C. Kuo, and M. A. Tsai, “Oblique electron-beam evaporation of distinctive indium-tin-oxide nanorods for enhanced light extraction from InGaN/GaN light emitting diodes,” Opt. Express 17(23), 21250–21256 (2009). [CrossRef] [PubMed]
22. J. Gao, R. Chen, D. H. Li, L. Jiang, J. C. Ye, X. C. Ma, X. D. Chen, Q. H. Xiong, H. D. Sun, and T. Wu, “UV light emitting transparent conducting tin-doped indium oxide (ITO) nanowires,” Nanotechnology 22(19), 195706 (2011). [CrossRef] [PubMed]
23. Y. F. Yao, C. G. Tu, T. W. Chang, H. T. Chen, C. M. Weng, C. Y. Su, C. Hsieh, C. H. Liao, Y. W. Kiang, and C. C. Yang, “Growth of highly conductive Ga-doped ZnO nanoneedles,” ACS Appl. Mater. Interfaces 7(19), 10525–10533 (2015). [CrossRef] [PubMed]
24. C. G. Tu, C. H. Liao, Y. F. Yao, H. S. Chen, C. H. Lin, C. Y. Su, P. Y. Shih, W. H. Chen, E. Zhu, Y. W. Kiang, and C. C. Yang, “Regularly patterned non-polar InGaN/GaN quantum-well nanorod light-emitting diode array,” Opt. Express 22(S7), A1799–A1809 (2014). [CrossRef] [PubMed]
25. Y. F. Yao, H. T. Chen, C. Y. Su, C. Hsieh, C. H. Lin, Y. W. Kiang, and C. C. Yang, “Phosphor-free, white-light LED under alternating-current operation,” Opt. Lett. 39(22), 6371–6374 (2014). [CrossRef] [PubMed]
26. Y. Yoshida, S. Tanaka, I. Hiromitsu, Y. Fujita, and K. Yoshino, “Ga-doped ZnO film as a transparent electrode for phthalocyanine/perylene heterojunction solar cell,” Jpn. J. Appl. Phys. 47(2), 867–871 (2008). [CrossRef]
27. W. Gu, T. Xu, and J. Zhang, “Improved ohmic contact of Ga-doped ZnO to p-GaN by using copper sulfide intermediate layers,” Solid-State Electron. 89, 76–80 (2013). [CrossRef]
28. Y. Kuo, S. Y. Ting, C. H. Liao, J. J. Huang, C. Y. Chen, C. Hsieh, Y. C. Lu, C. Y. Chen, K. C. Shen, C. F. Lu, D. M. Yeh, J. Y. Wang, W. H. Chuang, Y. W. Kiang, and C. C. Yang, “Surface plasmon coupling with radiating dipole for enhancing the emission efficiency of a light-emitting diode,” Opt. Express 19(S4), A914–A929 (2011). [CrossRef] [PubMed]
29. C. H. Lin, C. Hsieh, C. G. Tu, Y. Kuo, H. S. Chen, P. Y. Shih, C. H. Liao, Y. W. Kiang, C. C. Yang, C. H. Lai, G. R. He, J. H. Yeh, and T. C. Hsu, “Efficiency improvement of a vertical light-emitting diode through surface plasmon coupling and grating scattering,” Opt. Express 22(S3), A842–A856 (2014). [CrossRef] [PubMed]
30. C. H. Lin, C. Y. Su, Y. Kuo, C. H. Chen, Y. F. Yao, P. Y. Shih, H. S. Chen, C. Hsieh, Y. W. Kiang, and C. C. Yang, “Further reduction of efficiency droop effect by adding a lower-index dielectric interlayer in a surface plasmon coupled blue light-emitting diode with surface metal nanoparticles,” Appl. Phys. Lett. 105(10), 101106 (2014). [CrossRef]
31. C. H. Lin, C. H. Chen, Y. F. Yao, C. Y. Su, P. Y. Shih, H. S. Chen, C. Hsieh, Y. Kuo, Y. W. Kiang, and C. C. Yang, “Behaviors of surface plasmon coupled light-emitting diodes induced by surface Ag nanoparticles on dielectric interlayers,” Plasmonics 10(5), 1029–1040 (2015). [CrossRef]
32. C. F. Lu, C. H. Liao, C. Y. Chen, C. Hsieh, Y. W. Kiang, and C. C. Yang, “Reduction in the efficiency droop effect of a light-emitting diode through surface plasmon coupling,” Appl. Phys. Lett. 96(26), 261104 (2010). [CrossRef]
33. C. H. Lin, C. Y. Su, E. Zhu, Y. F. Yao, C. Hsieh, C. G. Tu, H. T. Chen, Y. W. Kiang, and C. C. Yang, “Modulation behaviors of surface plasmon coupled light-emitting diode,” Opt. Express 23(6), 8150–8161 (2015). [CrossRef] [PubMed]
34. D. M. Yeh, C. F. Huang, C. Y. Chen, Y. C. Lu, and C. C. Yang, “Localized surface plasmon-induced emission enhancement of a green light-emitting diode,” Nanotechnology 19(34), 345201 (2008). [CrossRef] [PubMed]
35. Y. Kuo, H. T. Chen, W. Y. Chang, H. S. Chen, C. C. Yang, and Y. W. Kiang, “Enhancements of the emission and light extraction of a radiating dipole coupled with localized surface plasmon induced on a surface metal nanoparticle in a light-emitting device,” Opt. Express 22(S1), A155–A166 (2014). [CrossRef] [PubMed]
36. M. Y. Choi, D. Choi, M. J. Jin, I. Kim, S. H. Kim, J. Y. Choi, S. Y. Lee, J. M. Kim, and S. W. Kim, “Mechanically powered transparent flexible charge-generating nanodevices with piezoelectric ZnO nanorods,” Adv. Mater. 21(21), 2185–2189 (2009). [CrossRef]
37. C. Y. Chen, C. Hsieh, C. H. Liao, W. L. Chung, H. T. Chen, W. Cao, W. M. Chang, H. S. Chen, Y. F. Yao, S. Y. Ting, Y. W. Kiang, C. C. Yang, and X. Hu, “Effects of overgrown p-layer on the emission characteristics of the InGaN/GaN quantum wells in a high-indium light-emitting diode,” Opt. Express 20(10), 11321–11335 (2012). [CrossRef] [PubMed]
38. L. W. Jang, D. W. Jeon, T. Sahoo, D. S. Jo, J. W. Ju, S. J. Lee, J. H. Baek, J. K. Yang, J. H. Song, A. Y. Polyakov, and I. H. Lee, “Localized surface plasmon enhanced quantum efficiency of InGaN/GaN quantum wells by Ag/SiO2 nanoparticles,” Opt. Express 20(3), 2116–2123 (2012). [CrossRef] [PubMed]
